On-device LLM inference for Android - 100% offline, GPU-accelerated via llama.cpp + Vulkan
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).